Advanced Robot Assistance System Using Depth Camera and AI

This paper presents a novel system combining real-time collision warning, pedestrian warning, cyclist warning and slope detection using Intel Real Sense D455 and Object detection module. The system detects the inclination angle with respect to ground of an Unmanned Ground Vehicle(UGV), alerts the UGV of potential collisions within a specified range and provides warnings if a pedestrian and cyclist is detected. The proposed approach integrates multiple sensing modalities and machine learning to enhance the safety and awareness of UGV, particularly in environments with complex terrain and condition.